BC108 Fuzz Face Oscillation Problem

[maddmann08]

I have just entered the world of DIY pedal building with a fuzz face clone, and this is my first post here so yay! I used a pbc purchased from General Guitar Gadgets. I based it on their standard NPN Silicon version (http://www.generalguitargadgets.com/projects/73-fuzz-tones/101-fuzz-faces) and added an external bias pot and pregain control. I used all values from the Axis Face on Fuzz Central (http://fuzzcentral.ssguitar.com/axisface.php) and BC108b transistors from Smallbear. The pedal sounds incredible, the only issue is that when the fuzz and volume pot are both maxed out, I get a very high pitched oscillation. The problem disappears when the volume on the pedal is turned down to about 11 o'clock, but it's always more fun to run a fuzz at full tilt. After searching the forums here I saw that it could potentially be a problem with the wiring layoutand cross-talk between wires, but as this is my first build I'm not really sure. Here's a picture of the inside as best I could get it:



Any comments/tips on my wiring are definitely welcome, as I am trying to learn! Also, if the wiring doesn't look to be the problem a point in the right direction would also be appreciated.

[Jaicen_solo]

The answer is that you probably have very high gain transistors. A really simple solution is to put a small resistance in series with the fuzz pot. Start at 100 Ohms and see if that helps.

[GGBB]

In my BC108C fuzz I found a 4700pF cap between Q1 collector and base to be quite effective.

[chromesphere]

Its definitely the battery.  Do not use 4site batteries whatever you do ha.

Seriously, the 100 ohm trick might work.  You could use a temporary 1k pot and adjust it till the oscillation goes away and find the right value.  So you don't lose any more gain then you have to.  This is a trick I've seen people do before on tonebenders for the same reason.

Paul

[Jaicen_solo]

A 1k pot will be hard to make small enough. You only need something just a bit more than zero resistance.

[chromesphere]

100 ohm pot then?

[maddmann08]

Thanks for the tips, I will try these out when I get the chance. In the mean time you bring up another point that I've been curious about; how much do batteries actually matter in these circuits?

[maddmann08]

Also, what kind of effect on the tone will the above suggestions have?

[R.G.]

I can't tell from the pics - is there a power supply bypass of at least 22uF electrolytic and 0.1uF ceramic between power and ground on the PCB?

[LucifersTrip]

before you make any adjustments, it'd be cool to make sure your voltages are in the ballpark

if you use a cap between Q1 B & C, only a 50 - 200pf should be necessary

[GGBB]

before you make any adjustments, it'd be cool to make sure your voltages are in the ballpark

if you use a cap between Q1 B & C, only a 50 - 200pf should be necessary

Could be the norm but in my case that wasn't sufficient.  I tried many values from 10pF to 1000pF and 470pf (I typoed before when I wrote 4700) was the lowest value that I was able to get away with.  Possibly high gain transistors (hFE > 500) or the circuit particulars itself make a difference.

[maddmann08]

I can't tell from the pics - is there a power supply bypass of at least 22uF electrolytic and 0.1uF ceramic between power and ground on the PCB?

I believe so, there is a 22 uF electrolytic between the fuzz pot and ground, and a .1 uF film type before the volume pot. Is that what you mean? here's the layout I used:
http://www.generalguitargadgets.com/pdf/ggg_ff5_lo_npn.pdf?phpMyAdmin=78482479fd7e7fc3768044a841b3e85a

*on taking a second look, the electrolytic off the fuzz pot is 15 uF instead of 22, as this is the value given on Fuzz Central for the Axis fuzz. Could this be part of the problem?

this shows a .01 uF cap, but I did use a .1 because that's what all other schematics showed.

before you make any adjustments, it'd be cool to make sure your voltages are in the ballpark

if you use a cap between Q1 B & C, only a 50 - 200pf should be necessary

When biased normally, the transistor pin voltages are:
Q1
B- 0.61
E- 0.0
C- 1.29
Q2
B- 1.29
E- 0.70
C- 4.86

I think these numbers look good from what I've read. The oscillation also stops when I starve Q2 to under one volt on the collector (the other voltages remain relatively unchanged), but that is to be expected since it is not amplifying the signal to the same degree.

[LucifersTrip]

Voltages are good, so I'd now go with some combo of the good suggestions above. BC cap on Q1 and/or Q2 (start 50pf and work your way up) and/or limiter trim on fuzz pot.

GGG

Q1
Collector
1.4v
Base
0.6v
Emitter
0.0v

Q2
Collector
4.5v
Base
1.4v
Emitter
0.8v

[maddmann08]

Thanks for all the help, I'll try these as soon as I can get my hands on the parts. Are these suggestions fairly transparent or should I expect some change in sound (loss of top end, gain, etc.)?

[LucifersTrip]

Thanks for all the help, I'll try these as soon as I can get my hands on the parts. Are these suggestions fairly transparent or should I expect some change in sound (loss of top end, gain, etc.)?

BC cap will tame it a bit and have some loss of high end....trimmer on fuzz will do the obvious...stop you from hitting 100% fuzz.
that is the reason why, in both cases, you start small and work your way up

[R.G.]

I believe so, there is a 22 uF electrolytic between the fuzz pot and ground, and a .1 uF film type before the volume pot. Is that what you mean? here's the layout I used:
*on taking a second look, the electrolytic off the fuzz pot is 15 uF instead of 22, as this is the value given on Fuzz Central for the Axis fuzz. Could this be part of the problem?

this shows a .01 uF cap, but I did use a .1 because that's what all other schematics showed.

Those are not power supply filter caps. They're used for signal purposes in the rest of the circuit. Oscillation in circuits that work with low gain (i.e. germanium amoung others) transistors but oscillate with high gain silicon is often a power supply or grounding issue. Perhaps not, but it's always a good guess.

[maddmann08]

I'm running the pedal completely on battery power and do not plan on adding a dc jack as I have heard this can introduce a lot of noise to the circuit. I assume this would make power supply filtering unnecessary? Also I have considered the possibility of a grounding issue, but I have not been able to find much good info on proper grounding for effect pedals. Would you be able to point me in the right direction?

[R.G.]

I'm running the pedal completely on battery power and do not plan on adding a dc jack as I have heard this can introduce a lot of noise to the circuit.

"I have heard" is most often an insufficient reason for doing anything technical.

And in this case, not adding a DC jack because of a perceived noise issue is somewhere between incomplete and wrong. Many similar if not almost identical devices run from a DC jack. Takes some doing if you use PNP and positive ground, but with BC108s in there, that's probably not the case. It *is* possible to find a bad/noisy DC adapter, but there is no inherent reason to think this can't have a perfectly acceptable outcome.

I assume this would make power supply filtering unnecessary?

No, that is incorrect. It is entirely possible to power things from batteries and have them oscillate unintentionally. The problem lies in there being some length of wire between the battery and circuit, and that wire having a non-zero resistance and inductance. Decoupling caps need to be as close to the power usage point as it's physically possible to get them. In this case, putting them on the PCB would be a Very Good Idea.

What makes this an issue is the very high bandwidth and gain of those silicon transistors. That can cause problems (and advantages, in other circuits) that germanium either doesn't have (or can't match).

Also I have considered the possibility of a grounding issue, but I have not been able to find much good info on proper grounding for effect pedals. Would you be able to point me in the right direction?

I regularly post comments on proper grounding here. The problem with giving a cookbook is that in pedals, unless there is a true mistake in grounding, it's easy to get a way with an error that would be fatal in a more complex setup. For instance, in that wiring diagram, even if you did it identical to the diagram, 100% of the power for the circuit travels down the wire carrying the input signal ground, so any currents in that wire cause a voltage to appear across the wire because the wire is not a perfect conductor. It has some resistance, however small. And that signal is added to the input.

With high enough gain, that is enough signal to make the thing oscillate. It may not be THE cause in this situation, but it's a possibility. This wiring won't go away in pedals because many pedals are OK with it - until you get one that's not.

[maddmann08]

I assume this would make power supply filtering unnecessary?

No, that is incorrect. It is entirely possible to power things from batteries and have them oscillate unintentionally. The problem lies in there being some length of wire between the battery and circuit, and that wire having a non-zero resistance and inductance. Decoupling caps need to be as close to the power usage point as it's physically possible to get them. In this case, putting them on the PCB would be a Very Good Idea

Ahh I see. That clears it up; on GGG the build info said those power supply filtering caps were only necessary if using a DC jack. I guess it pays to check a few different sources of information before a build! As soon as I have the time to tinker with it again I'll give this a shot. At least I have a plan of attack now.

And thank you all for your advice and patience. I have a lot to learn but it seems like the start to a worthwhile journey, I don't think I'll be able to stop after just one...